One important element for evaluating the performance of an analyzing device is data reproducibility. A value frequently used as an index which quantitatively represents the reproducibility is a CV (coefficient of variation) value, which is the standard deviation (STDEV) divided by the average (Average) and multiplied by 100 (the CV value may also be called a relative standard deviation). For example, when a mass spectrometry is performed using a liquid chromatograph mass spectrometer (LC/MS), a calibration curve needs to be previously created based on the result of a measurement of a sample having a known concentration, in which the CV value is extremely important in determining the calibration points. According to a standard set by the U.S. Food and Drug Administration (FDA), the CV value should be 20% or lower at the calibration point corresponding to the limit of quantitation (LOQ) and 15% or lower at the other points.
In general, the CV value of an LC/MS or another type of chromatograph system is calculated from the area values of a plurality of chromatogram peaks (see Patent Literature 1 or other references). However, a CV value calculated based on a chromatogram obtained through one measurement normally does not only contain statistical dispersion factors for the value but also other unstable factors depending on the measurement conditions or the like. Whether or not such unstable factors due to the measurement conditions or the like can be removed so as to acquire stable and highly-reproducible data is critical in evaluating the device as well as in improving the accuracy of the quantitative analysis.
The calculation of the CV value from the area values of chromatogram peaks can be easily achieved using generally available analyzing software products. However, those products are not capable of isolating or identifying unstable factors due to the measurement conditions or the like contained in the calculated CV value. Therefore, it has been difficult for analysis operators to determine whether or not the device is delivering adequately high performance, or whether or not the measurement conditions are appropriate, based on the CV value obtained as a result of an actual measurement.